GEO 352P Final Project - National Park Visitors and Biodiversity¶

Tianyi Xu 11/4/2022¶

Introduction¶

Yosemite Valley, Thunderstorm by Ansel Adams

 

 

 

 

 

 

 

 

 

 

 

 

The first national park, Yellowstone, was established in Montana and Wyoming as a public park for visitors in 1872. Later in 1916, President Woodrow Wilson started the National Park Service, which was responsible for protecting the 35 national parks and monuments then managed by the department. As of 2022, there are 63 national parks proteced by the National Park Service across 30 states and two US territories. With scenic drives, wildlife and flora, hiking trails, historic sites, these national parks are enjoyed by countless domestic and international visitors every year who seek for amazing adventures in natural wonders.

Alt Text: Yosemite Valley, Thunderstorm by Ansel Adams

Source: https://www.anseladams.com/thunderstorm-yosemite-valley/

Methods¶

Packages¶

For this project, I am using numpy and pandas for data cleaning and analysis, and matplotlib, seaborn, plotly, and cartopy for making figures.

In [1]:
# Import packages
import requests
import json
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np
from netCDF4 import Dataset as netcdf_dataset
import cartopy
import cartopy.feature as cfeature
from cartopy.mpl.ticker import LongitudeFormatter, LatitudeFormatter
import cartopy.crs as ccrs
import warnings
warnings.filterwarnings("ignore") 
import cartopy.io.img_tiles as cimgt
import seaborn as sns
from matplotlib.colors import ListedColormap
import plotly.graph_objects as go
import plotly.io as pio
import plotly.express as px
pio.renderers.default = "notebook"

Color Palette¶

For all the figures produced in this project, I am using a color palette inspired from Acadia National Park, one of my favorite national parks! This palette is also colorblind-friendly. For larger barplots, I am using the seaborn palete magma.

Source: https://github.com/kevinsblake/NatParksPalettes

Yosemite Valley, Thunderstorm by Ansel Adams
In [2]:
# Create a palette using hex values
acadia = ['#212e51','#434e7e', '#8087aa', '#b7abbc', '#f9ece8', '#fcc893', '#feb424', '#fd8700', '#d8511c']
acadia7 = ['#212e51','#434e7e', '#8087aa', '#b7abbc', '#f9ece8', '#fcc893', '#feb424']
sns.palplot(sns.color_palette(acadia))

Data¶

Datasets used in this project are publicly available on Kaggle and Data World.

Sources:

  • https://www.kaggle.com/datasets/aliamini587/biodiversity-in-national-parks
  • https://data.world/inform8n/us-national-parks-visitation-1904-2016-with-boundaries
In [3]:
# Import datasets
parks = pd.read_csv('parks.csv')
species = pd.read_csv('species.csv')
visit = pd.read_csv('allvisit.csv')

# Extract only information of interest and cleaning data
visit = visit[visit['Unit Type'] == "National Park"]
visit = visit.rename(columns = {'Unit Name': 'Park'})
visit = visit[['Region', 'State', 'Unit Code', 'Park', 'Visitors', 'YearRaw']]
visit.drop(visit.loc[visit['Park'] == "Denali National Preserve"].index, inplace = True)

Parks is a dataset with park names, sizes, and geographic locations.

In [4]:
parks = parks.rename(columns = {'Park Name': 'Park'})
parks.head(3)
Out[4]:
Park Code Park State Acres Latitude Longitude
0 ACAD Acadia National Park ME 47390 44.35 -68.21
1 ARCH Arches National Park UT 76519 38.68 -109.57
2 BADL Badlands National Park SD 242756 43.75 -102.50
In [5]:
# Change acres column to float type
parks['Acres'] = parks['Acres'].astype(float)/1000
In [6]:
size = parks['Acres']
size.dtype
Out[6]:
dtype('float64')

Species is a dataset with park names and information of all species observations recorded in the parks.

In [7]:
species = species.rename(columns = {'Park Name': 'Park'})
species.head(3)
Out[7]:
Species ID Park Category Order Family Scientific Name Common Names Record Status Occurrence Nativeness Abundance Seasonality Conservation Status
0 ACAD-1000 Acadia National Park Mammal Artiodactyla Cervidae Alces alces Moose Approved Present Native Rare Resident NaN
1 ACAD-1001 Acadia National Park Mammal Artiodactyla Cervidae Odocoileus virginianus Northern White-Tailed Deer, Virginia Deer, Whi... Approved Present Native Abundant NaN NaN
2 ACAD-1002 Acadia National Park Mammal Carnivora Canidae Canis latrans Coyote, Eastern Coyote Approved Present Not Native Common NaN Species of Concern

Visit is a dataset with park names and annual visitor counts from 1904 to 2016.

In [8]:
visit.head(3)
Out[8]:
Region State Unit Code Park Visitors YearRaw
0 PW OR CRLA Crater Lake National Park 1500.0 1904
3 PW WA OLYM Olympic National Park 2200.0 1935
5 NE ME ACAD Acadia National Park 64000.0 1919

Data Analysis and Visualization¶

Display National Park Distribution Using Cartopy¶

In [9]:
# Set up the figure and use platecarree projection
plt.figure(figsize = (15, 15))
ax = plt.axes(projection = ccrs.PlateCarree())

# Create and set extent for the map
extent = [-179, -50, 10, 55]
ax.set_extent(extent)

# Add features to the map
# ax.add_feature(cartopy.feature.OCEAN)
ax.add_feature(cartopy.feature.LAND, color = '#b7abbc', alpha = 0.5)
ax.add_feature(cartopy.feature.LAKES, color = 'white')
ax.add_feature(cartopy.feature.RIVERS, color = 'white')
ax.axis('off')

# Add gridlines
ax.gridlines(color = 'white', alpha = 0.5)

# You can also add stamen terrain background to the map
# stamen_terrain = cimgt.Stamen('terrain-background')
# ax.add_image(stamen_terrain, 4)

# Plot national park locations
plt.scatter(parks['Longitude'], parks['Latitude'], color = '#212e51', alpha = 0.5, s = size/10,
         transform = ccrs.PlateCarree(), label = 'National Parks')
plt.legend(loc = 'lower right', fontsize = 13)
plt.title("US National Park Distribution", fontsize = 14);

Time-series Analysis with National Park Visitors from 1904-2016 and Total Visitors¶

In [10]:
# Unique states
unistate = visit["State"].unique()
print(unistate)

['OR' 'WA' 'ME' 'TX' 'FL' 'NC' 'MI' 'CA' 'VA' 'CO' 'UT' 'NM' 'WY' 'HI'
 'KY' 'NV' 'ND' 'VI' 'AZ' 'SD' 'OH' 'SC' 'AR' 'AK' 'AS' 'MT' 'MN']
In [11]:
# Unique years
print(visit["YearRaw"].unique())
print(visit["YearRaw"].unique().shape)

['1904' '1935' '1919' '1944' '1948' '1931' '1940' '1971' '1936' '1934'
 '1929' '1937' '1924' '1921' '1906' '1908' '1963' '1965' '1972' '1960'
 '1938' '1941' '1923' '1957' '1915' '1933' '1917' '1939' '1978' '1985'
 '1920' '1922' '1982' '1956' '1970' '1932' '2002' '1911' '1976' 'Total'
 '2016' '2015' '2014' '2013' '2012' '2011' '2010' '2009' '2008' '2007'
 '2006' '2005' '2003' '2004' '2001' '2000' '1999' '1998' '1997' '1996'
 '1995' '1994' '1993' '1992' '1991' '1990' '1989' '1988' '1987' '1986'
 '1984' '1983' '1981' '1980' '1979' '1977' '1975' '1974' '1973' '1969'
 '1968' '1967' '1966' '1964' '1962' '1961' '1959' '1958' '1955' '1954'
 '1953' '1952' '1951' '1950' '1949' '1947' '1946' '1945' '1943' '1942'
 '1930' '1928' '1927' '1926' '1925' '1918' '1916' '1914' '1913' '1912'
 '1910' '1909' '1905' '1907']
(114,)

Notice that there is a value of total visitors of all years, which would make the analysis harder. Here, I made a new dataset just for total visitors of all national parks, and deleted rows with total visitors on the original dataset.

In [12]:
# Create a new dataset only with total visitors
totalvisit = visit[visit['YearRaw'] == 'Total']

# Sorting the dataset by visitor numbers
totalvisit.sort_values(by = ['Visitors'], ascending = False).head()
Out[12]:
Region State Unit Code Park Visitors YearRaw
412 SE NC GRSM Great Smoky Mountains National Park 521947058.0 Total
652 IM AZ GRCA Grand Canyon National Park 205486894.0 Total
708 PW CA YOSE Yosemite National Park 190404561.0 Total
644 IM CO ROMO Rocky Mountain National Park 180076609.0 Total
496 IM WY YELL Yellowstone National Park 175040126.0 Total
In [13]:
totalvisit.sort_values(by = ['Visitors']).head()
Out[13]:
Region State Unit Code Park Visitors YearRaw
760 PW AS NPSA National Park of American Samoa 113694.0 Total
743 AK AK KOVA Kobuk Valley National Park 156655.0 Total
735 AK AK GAAR Gates of the Arctic National Park 225813.0 Total
744 AK AK LACL Lake Clark National Park 363894.0 Total
414 MW MI ISRO Isle Royale National Park 1071614.0 Total

Looks like the most visited national parks are Great Smoky, Grand Canyon, Yosemite, Rocky Mountain, and Yellowstone. Not surprising at all! I've been to 4/5. Maybe it's time for us to visit some less well-known national parks, such as parks in the Pacific Ocean and Alaska.

Below, I am plotting visitor counts in states with the most national parks: California, Alaska, and Utah.

In [29]:
# National parks in California
ca = totalvisit[totalvisit['State'] == 'CA']

# Barplot
sns.set(font_scale = 1)
sns.set_style("white")
plt.figure(figsize = (10, 6))

sns.barplot(data = ca, x = "Visitors", y = "Park",
            palette = acadia, order = ca.sort_values('Visitors', ascending = False).Park).set(title = 'Total National Park Visitors in California')
sns.despine();
In [30]:
# National parks in Alaska
ak = totalvisit[totalvisit['State'] == 'AK']

# Barplot
sns.set(font_scale = 1)
sns.set_style("white")
plt.figure(figsize = (10, 6))

sns.barplot(data = ak, x = "Visitors", y = "Park",
            palette = acadia, order = ak.sort_values('Visitors', ascending = False).Park).set(title = 'Total National Park Visitors in Alaska')
sns.despine();
In [31]:
# National parks in Utah
ut = totalvisit[totalvisit['State'] == 'UT']

# Barplot
sns.set(font_scale = 1)
sns.set_style("white")
plt.figure(figsize = (10, 6))

sns.barplot(data = ut, x = "Visitors", y = "Park",
            palette = acadia, order = ut.sort_values('Visitors', ascending = False).Park).set(title = 'Total National Park Visitors in Utah')
sns.despine();

Which states have the most national park visitors? The answer is North Carolina, with the Great Smokies!

In [20]:
# Group national park visitors by state and sort the value
state = pd.DataFrame(totalvisit.groupby('State')['Visitors'].sum(), columns = ['Visitors'])
state = state.reset_index()
state.sort_values(by = ['Visitors'], ascending = False, inplace = True)

# Barplot
plt.figure(figsize = (12, 6))
sns.set(font_scale = 1)
sns.set_style("white")
sns.barplot(data = state, x = "Visitors", y = "State",
            palette = acadia, order = state.sort_values('Visitors', ascending = False).State).set(title = 'Total National Park Visitors across States')
sns.despine();
In [21]:
# Drop total visitor rows from the visitor dataset
visit.drop(visit.loc[visit['YearRaw'] == "Total"].index, inplace = True)
visit.sort_values(by = ['YearRaw'], inplace = True)
print(visit.shape)
visit.head()

(4587, 6)
Out[21]:
Region State Unit Code Park Visitors YearRaw
0 PW OR CRLA Crater Lake National Park 1500.0 1904
345 MW AR HOSP Hot Springs National Park 101000.0 1904
339 MW SD WICA Wind Cave National Park 2900.0 1904
315 PW WA MORA Mount Rainier National Park 563.0 1904
206 PW CA KICA Kings Canyon National Park 1000.0 1904

After some further data cleaning, we are ready for plotting a time-series graph of national park visitors through time.

In [22]:
# Visitors through time of all national parks
plt.figure(figsize = (12, 6))
sns.set_style('white')
sns.lineplot(data = visit, x = 'YearRaw', y = 'Visitors', lw = 3, color = '#feb424').set(title = 'National Park Visitors through Time')
plt.xticks(np.arange(0, len(set(visit['YearRaw'])) + 1, 10))

sns.despine();

What about the increasing trend of national park visitors for individual state? Here, I used plotly to create an interactive plot.

In [23]:
# Group visitors by states and years for visualization
stateyear = pd.DataFrame(visit.groupby(['YearRaw', 'State'])['Visitors'].sum(), columns = ['Visitors'])
stateyear = stateyear.reset_index()
stateyear.head()
Out[23]:
YearRaw State Visitors
0 1904 AR 101000.0
1 1904 CA 1000.0
2 1904 OR 1500.0
3 1904 SD 2900.0
4 1904 WA 563.0
In [24]:
# Extract the magma palette from sns
pal = list(sns.color_palette(palette = 'magma', n_colors = 27).as_hex())

# Set up the template
pio.templates.default = "simple_white"
fig = go.Figure()

# Plotly graph
for state, color in zip(unistate, pal):
    fig.add_trace(go.Scatter(x = stateyear[stateyear['State'] == state]['YearRaw'],
                             y = stateyear[stateyear['State'] == state]['Visitors'],
                             name = state,
                             line_color = color, 
                             fill = None)) 
# Add layout
fig.update_layout(title = dict(text = "National Park Visitors through Time Across States"), 
                  xaxis = dict(title = "Year"), 
                  yaxis = dict(title = "Visitors"))

# Add two buttons: select and deselect all
fig.update_layout(dict(updatemenus = [dict(type = "buttons", direction = "left",
                                         buttons = list([dict(args = ["visible", "legendonly"], label = "Deselect All", method = "restyle"),
                                                       dict(args = ["visible", True], label = "Select All", method = "restyle")]),
                                         pad = {"r": 10, "t": 10},
                                         showactive = False, x = 1, xanchor = "right",
                                         y = 1.1, yanchor = "top")]))
fig.show()

Most of the national parks have an increasing number of visitors through time, but a minority of them have a relatively stable number.

National Park Biodiversity Analysis¶

What species are found in national parks across the US?

In [25]:
# Unique species category
print(species["Category"].unique())

['Mammal' 'Bird' 'Reptile' 'Amphibian' 'Fish' 'Vascular Plant'
 'Spider/Scorpion' 'Insect' 'Invertebrate' 'Fungi' 'Nonvascular Plant'
 'Crab/Lobster/Shrimp' 'Slug/Snail' 'Algae']
In [26]:
# Numbers of species for each category
species['Category'].value_counts()
Out[26]:
Vascular Plant         65221
Bird                   14601
Insect                 14349
Fungi                   6203
Nonvascular Plant       4278
Fish                    3956
Mammal                  3867
Invertebrate            1566
Reptile                 1343
Algae                    976
Slug/Snail               787
Spider/Scorpion          776
Amphibian                743
Crab/Lobster/Shrimp      582
Name: Category, dtype: int64
In [33]:
# Create a dataset with category counts
category = pd.DataFrame(species.groupby("Category").size(), columns = ["Count"])
category = category.reset_index()

# Plot taxon count for each category
sns.set(font_scale = 1)
sns.set_style("white")
plt.figure(figsize = (10, 6))

sns.barplot(data = category, x = "Count", y = "Category", 
            order = category.sort_values('Count', ascending = False).Category,
           palette = acadia).set(title = 'Taxon Count per Category')
sns.despine();

The majority of species found in national parks are vascular plant, which is not surprised at all! However, I am a vertebrate biologist.

In [34]:
# Select only vertebrates from the biodiversity data
vert = species[species['Category'].isin(['Bird','Fish', 'Mammal', 'Reptile', 'Amphibian'])]
# Species count in each category
vert_category = pd.DataFrame(vert.groupby("Category").size(), columns = ["Count"])
vert_category = vert_category.reset_index()
vert_category
Out[34]:
Category Count
0 Amphibian 743
1 Bird 14601
2 Fish 3956
3 Mammal 3867
4 Reptile 1343
In [38]:
# Barplot with only vertebrate species count
sns.set(font_scale = 1)
sns.set_style("white")
plt.figure(figsize = (10, 6))

sns.barplot(data = vert_category, x = "Count", y = "Category",
           palette = acadia, 
           order = vert_category.sort_values('Count', ascending = False).Category).set(title = 'Vertebrate Species Count')
sns.despine();

There are a lot of birds, but not many herps. Let's only consider terrestrial vertebrates (non-fish vertebrates).

In [39]:
# Select non-fish vertebrates
nfvert = species[species['Category'].isin(['Bird', 'Mammal', 'Reptile', 'Amphibian'])]
nfvert_category = pd.DataFrame(nfvert.groupby("Category").size(), columns = ["Count"])
nfvert_category = nfvert_category.reset_index()
nfvert_category
Out[39]:
Category Count
0 Amphibian 743
1 Bird 14601
2 Mammal 3867
3 Reptile 1343
In [40]:
# Group and sort dataset by vertebrate biodiversity
nfvertdiversity = nfvert.groupby("Park")["Scientific Name"].count().reset_index(name = 'Count')
nfvertdiversity.sort_values(by = ['Count'], ascending = False).head()
Out[40]:
Park Count
14 Death Valley National Park 904
44 Redwood National Park 727
21 Grand Canyon National Park 653
3 Big Bend National Park 579
9 Carlsbad Caverns National Park 536

The top 5 national parks with highest non-fish vertebrate biodiversity are Death Valley, Redwood, Grand Canyon, Big Bend, and Carlsbad Caverns.

In [41]:
# Add biodiversity data to the parks dataset for location information
parkdiversity = pd.merge(parks, nfvertdiversity)
parkdiversity.head(3)
Out[41]:
Park Code Park State Acres Latitude Longitude Count
0 ACAD Acadia National Park ME 47.390 44.35 -68.21 445
1 ARCH Arches National Park UT 76.519 38.68 -109.57 292
2 BADL Badlands National Park SD 242.756 43.75 -102.50 326

With a merged dataset, we are able to plot vertebrate biodiversity data on the US map.

In [42]:
plt.figure(figsize = (15, 15))
ax = plt.axes(projection = ccrs.PlateCarree())

# Create and set extent for the map
extent = [-179, -50, 10, 55]
ax.set_extent(extent)

# Add features to the map
# ax.add_feature(cartopy.feature.OCEAN)
ax.add_feature(cartopy.feature.LAND, color = '#b7abbc', alpha = 0.5)
ax.add_feature(cartopy.feature.LAKES, color = 'white')
ax.add_feature(cartopy.feature.RIVERS, color = 'white')
ax.axis('off')

# Add gridlines
ax.gridlines(color = 'white', alpha = 0.5)

# Plot national park locations
plt.scatter(parks['Longitude'], parks['Latitude'], color = "#fd8700", alpha = 0.3, s = parkdiversity["Count"]*2,
         transform = ccrs.PlateCarree(), label = 'National Parks')
plt.legend(loc = 'lower right', fontsize = 13)
plt.title("US National Park Terrestrial Vertebrate Biodiversity", fontsize = 14);

What are the national parks with the most terrestrial vertebrate species of conservation concern?

In [43]:
# Count of conservation category in each park
conserve = nfvert[['Park', 'Conservation Status']].groupby(['Park', 'Conservation Status']).size()

# Convert to dataframe
conserve_df = conserve.to_frame().reset_index()
conserve_df.columns = ['Park', 'Conservation Status', 'Count']
In [50]:
iucn = ['Endangered', 'In Recovery', 'Species of Concern', 'Threatened']
plt.figure(figsize = (10, 6))

# Use a for loop to create barplots for each conservation category
for i in iucn:
    subset = conserve_df[conserve_df['Conservation Status'] == i]
    subset.sort_values(by = 'Count', ascending = False).plot(x = 'Park', y = 'Count', kind = 'bar', legend = None, color = acadia)
    plt.ylabel('Absolute number of species')
    plt.title('Conservation category: %s' % (i))
plt.show();

<Figure size 1000x600 with 0 Axes>

Let's take a closer look at reptile species! (I study them :D)

In [52]:
# Select only reptile, and use groupby to see reptile diversity in each park
reptile = vert[species['Category'] == 'Reptile']
reptilediversity = reptile.groupby("Park")["Scientific Name"].count().reset_index(name = 'Count')
reptilediversity.sort_values(by = ['Count'], ascending = False).head()
Out[52]:
Park Count
14 Death Valley National Park 90
19 Grand Canyon National Park 76
3 Big Bend National Park 62
9 Carlsbad Caverns National Park 62
40 Saguaro National Park 62

The parks with the most reptile species are Death Valley, Grand Canyon, Big Bend, Carlsbad Caverns, and Saguaro. Very interesting! I have been to 3/5, and will visit Big Bend and Carlsbad Caverns this Thanksgiving.

In [53]:
# The most common reptiles found in the park
famreptile = reptile.groupby("Family")["Scientific Name"].count().reset_index(name = 'Count')
famreptile.sort_values(by = ['Count'], ascending = False).head()
Out[53]:
Family Count
7 Colubridae 559
15 Iguanidae 246
25 Viperidae 99
11 Emydidae 90
19 Scincidae 67

Which national parks have invasive reptile species?

In [54]:
# Select the non-native reptile species
invasive = reptile[reptile["Nativeness"] == "Not Native"]
invasivecount = invasive.groupby("Park")["Scientific Name"].count().reset_index(name = 'Count')
invasivecount.sort_values(by = ['Count'], ascending = False, inplace = True)
In [56]:
sns.set(font_scale = 1)
sns.set_style("white")
plt.figure(figsize = (10, 6))

# Barplot of invasive species of reptiles in each park
sns.barplot(data = invasivecount, x = "Count", y = "Park",
           palette = acadia).set(title = 'Invasive Reptile Species Count')
sns.despine();

According to this barplot, national parks in Hawaii and Florida have the most invasive reptile species.

Explore Correlations between Variables¶

In this section, I hope to answer three questions with existing data.

Are there correlations between:

  • Terrestrial vertebrate species diversity and national park size?
  • Visitor counts and national park size?
  • Terrestrial vertebrate species diversity and visitor counts?
In [57]:
# Merge the biodiversity and total visit datasets
merge = pd.merge(parkdiversity, totalvisit, on = ['Park'])
merge.head()
Out[57]:
Park Code Park State_x Acres Latitude Longitude Count Region State_y Unit Code Visitors YearRaw
0 ACAD Acadia National Park ME 47.390 44.35 -68.21 445 NE ME ACAD 163512192.0 Total
1 ARCH Arches National Park UT 76.519 38.68 -109.57 292 IM UT ARCH 32577956.0 Total
2 BADL Badlands National Park SD 242.756 43.75 -102.50 326 MW SD BADL 69483454.0 Total
3 BIBE Big Bend National Park TX 801.163 29.25 -103.25 579 IM TX BIBE 15820237.0 Total
4 BISC Biscayne National Park FL 172.924 25.65 -80.08 309 SE FL BISC 18615078.0 Total
In [58]:
# Set a template for subplots
fig, axes = plt.subplots(1, 3, figsize = (20, 5))
sns.set(font_scale = 1.2)
sns.set_style("white")

# First plot: Biodiversity vs. Park Size
g = sns.regplot(ax = axes[0], x = "Acres", y = "Count", data = parkdiversity, color = "#8087aa")
g.set_title('Non-fish Vertebrate Biodiversity vs. National Park Size')
g.set(xscale = "log")

# Second plot: Visitors vs. Park Size
h = sns.regplot(ax = axes[1], x = "Acres", y = "Visitors", data = merge, color = "#434e7e")
h.set_title('Visitors vs. National Park Size')
h.set(xscale = "log")

# Third plot: Visitors vs. Biodiversity
j = sns.regplot(ax = axes[2], x = "Count", y = "Visitors", data = merge, color = "#212e51")
j.set_title('Visitors vs. Non-fish Vertebrate Biodiversity')

sns.despine();

According to the linear regression plots, there is no significant correlation between national park size and either biodiversity or visitor counts. This is probably because national park size varies to a great extent, and there are other factors playing a more important role in determining visitor counts and biodiversity. The number of visitors may be greatly impacted by parks' locations and reputation. Park biodiversity, on the other hand, may be related to climate and ecosystem in which the park is located.

What is interesting is the positive correlation between visitor counts and terrestrial vertebrate biodiversity. As the park biodiversity increases, the park tends to be more visitors. This is expected because many visitors go to national parks for their abundance of unique wildlife and ecosystems, like myself!

Future Directions¶

  • Effect of COVID-19 on visitor numbers of national parks
  • Addition of finer geographic data for national park shapes